From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:54793)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <ebotcazou@adacore.com>) id 1WXSpx-0008Em-CT
	for qemu-devel@nongnu.org; Tue, 08 Apr 2014 06:01:10 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <ebotcazou@adacore.com>) id 1WXSpo-00028b-1g
	for qemu-devel@nongnu.org; Tue, 08 Apr 2014 06:01:01 -0400
Received: from mel.v6.act-europe.fr ([2a02:2ab8:224:1::a0a:d2]:54838
	helo=smtp.eu.adacore.com) by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <ebotcazou@adacore.com>) id 1WXSpn-00027T-S0
	for qemu-devel@nongnu.org; Tue, 08 Apr 2014 06:00:51 -0400
From: Eric Botcazou <ebotcazou@adacore.com>
Date: Tue, 08 Apr 2014 12:00:04 +0200
Message-ID: <1615584.cDvrPMzPCT@polaris>
In-Reply-To: <CAFEAcA-kthmFCgcM5U5z97Wxi4NJx7XMePWcqPD9_eWF6b0swg@mail.gmail.com>
References: <5047457.AHxDUeBxbN@polaris> <16061188.ySlPLF5O9B@polaris>
	<CAFEAcA-kthmFCgcM5U5z97Wxi4NJx7XMePWcqPD9_eWF6b0swg@mail.gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Subject: Re: [Qemu-devel] [PATCH] target-arm: return more meaningful exit
	status
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: QEMU Developers <qemu-devel@nongnu.org>

> The reason we only exit with zero here is because the
> semihosting API doesn't provide a way for the application
> to report a non-zero exit code. In this case you're
> distinguishing "application exit" from the other
> failure reasons, which seems OK, though. I'm just
> wary of patches trying to fix this issue because
> in the past they've generally ignored the fact that
> we're implementing a fixed API here and can't make
> random additions to it just because they happen to work
> with newlib or whatever.

OK, I see.

> Returning a negative exit code is bogus, though. You
> probably just want
>    if (args == ADP_Stopped_ApplicationExit) {
>        code = EXIT_SUCCESS;
>    } else {
>        code = EXIT_FAILURE;
>    }

The idea was to distinguish the supported arguments from the unknown ones 
(newlib only uses the listed two) but that would also work for me.

-- 
Eric Botcazou